{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "70f6ae33",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import pandas as pd\n",
    "import time\n",
    "import re\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "from scipy.stats import chi2_contingency"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "98e4b2bc",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('annotated_readmes.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "05678686",
   "metadata": {},
   "source": [
    "### A) Methods: version 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "fac4cc5b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def clean_methods(x):\n",
    "    x = x.split(')')[0]\n",
    "    if x==\"A\" or x==\"B\" or x==\"C\" or x==\"D\":\n",
    "        return x\n",
    "    else:\n",
    "        return np.nan\n",
    "\n",
    "df['prompt_1'] = df['prompt_1'].apply(lambda x: clean_methods(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "46597202",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "A     5.22\n",
       "B    38.26\n",
       "C    46.96\n",
       "D     9.57\n",
       "Name: prompt_1, dtype: float64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df.loc[df.year==2020]['prompt_1'].value_counts(normalize = True).sort_index()*100).round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "459e7aa7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "A    20.35\n",
       "B    41.59\n",
       "C    37.17\n",
       "D     0.88\n",
       "Name: prompt_1, dtype: float64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df.loc[df.year==2021]['prompt_1'].value_counts(normalize = True).sort_index()*100).round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "46060f90",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11.58137127488887 0.00066615799533416\n",
      "0.21048032931925034 0.6463911870811092\n",
      "2.405453493573752 0.12091308021828988\n",
      "8.702404357082512 0.0031779060304225476\n"
     ]
    }
   ],
   "source": [
    "for col in ['A','B','C','D']:\n",
    "    a = df.loc[df.year==2020]['prompt_1'].value_counts().sort_index()[col]\n",
    "    b = len(df.loc[df.year==2020]) - a\n",
    "\n",
    "    c = df.loc[df.year==2021]['prompt_1'].value_counts().sort_index()[col]\n",
    "    d = len(df.loc[df.year==2021]) - c\n",
    "    test = chi2_contingency([[a,b],[c,d]],correction=False)\n",
    "    print(test[0],test[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "32ca17ac",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import matplotlib as mpl\n",
    "\n",
    "mpl.rcParams['axes.spines.right'] = False\n",
    "mpl.rcParams['axes.spines.top'] = False\n",
    "import mpl_toolkits.axisartist as axisartist\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "e23692d7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 450x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(4.5, 3))\n",
    "\n",
    "plt.barh(range(4)[::-1], (-df.loc[df.year==2020]['prompt_1'].\\\n",
    "                          value_counts(normalize = True).sort_index()*100),height = 0.9,\n",
    "         alpha=0.4, label = \"2020\")\n",
    "\n",
    "\n",
    "\n",
    "plt.barh(range(4)[::-1], (df.loc[df.year==2021]['prompt_1'].\\\n",
    "                          value_counts(normalize = True).sort_index()*100),height = 0.9,\n",
    "         alpha = 0.4, label = \"2021\")\n",
    "\n",
    "plt.legend(loc = \"upper right\")\n",
    "\n",
    "plt.xticks([-40,-20,0,20,40],[\"40%\",\"20%\",\"0%\",\"20%\",\"40%\"])\n",
    "\n",
    "plt.yticks([0,1,2,3],[\"A: Descriptive statistics and data visualization\",\n",
    "                      \"B: Statistical modeling and inference\",\n",
    "                      \"C: Machine learning and prediction\",\n",
    "                      \"D: Causal inference and counterfactuals\"][::-1])\n",
    "\n",
    "plt.xlim([-60,60])\n",
    "\n",
    "ax = plt.gca()\n",
    "\n",
    "# Looping through each ytick label and setting alignment to left\n",
    "for label in ax.get_yticklabels():\n",
    "    label.set_horizontalalignment('left')\n",
    "plt.gca().tick_params(axis='y', which='major', pad=240)\n",
    "\n",
    "plt.xlabel(\"Percentage of projects\")\n",
    "\n",
    "plt.ylabel(\"Data analysis type\",labelpad=20)\n",
    "\n",
    "plt.savefig(\"exploratory.pdf\",  bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "a97a5aa8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 450x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(4.5, 3))\n",
    "\n",
    "plt.barh(np.array(range(4)[::-1])+0.2, (df.loc[df.year==2020]['prompt_1'].\\\n",
    "                          value_counts(normalize = True).sort_index()*100),height = 0.4,\n",
    "         alpha=0.4, label = \"2020\")\n",
    "\n",
    "\n",
    "\n",
    "plt.barh(np.array(range(4)[::-1])-0.2, (df.loc[df.year==2021]['prompt_1'].\\\n",
    "                          value_counts(normalize = True).sort_index()*100),height = 0.4,\n",
    "         alpha = 0.4, label = \"2021\")\n",
    "\n",
    "plt.legend(loc = \"upper right\")\n",
    "\n",
    "plt.xticks([0,20,40],[\"0%\",\"20%\",\"40%\"])\n",
    "\n",
    "plt.yticks([0,1,2,3],[\"A: Descriptive statistics and data visualization\",\n",
    "                      \"B: Statistical modeling and inference\",\n",
    "                      \"C: Machine learning and prediction\",\n",
    "                      \"D: Causal inference and counterfactuals\"][::-1])\n",
    "\n",
    "plt.xlim([0,60])\n",
    "\n",
    "ax = plt.gca()\n",
    "\n",
    "# Looping through each ytick label and setting alignment to left\n",
    "for label in ax.get_yticklabels():\n",
    "    label.set_horizontalalignment('left')\n",
    "plt.gca().tick_params(axis='y', which='major', pad=240)\n",
    "\n",
    "plt.xlabel(\"Percentage of projects\")\n",
    "\n",
    "plt.ylabel(\"Data analysis type\",labelpad=20)\n",
    "\n",
    "plt.savefig(\"exploratory.pdf\",  bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b7f24399",
   "metadata": {},
   "source": [
    "### Check: Dataset type"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "ea8a620d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def process_data_type(x):\n",
    "    if x[0] in [\"1\",\"2\",\"3\",\"4\"]:\n",
    "        return x[0]\n",
    "    else:\n",
    "        print(x)\n",
    "        return np.nan\n",
    "\n",
    "df['prompt_2'] = df['prompt_2'].apply(lambda x: process_data_type(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "f606e28f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    63.48\n",
       "2    16.52\n",
       "3    14.78\n",
       "4     5.22\n",
       "Name: prompt_2, dtype: float64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df.loc[df['year']==2020]['prompt_2'].value_counts(normalize = True).sort_index()*100).round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "d4411e5a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    12.28\n",
       "2     2.63\n",
       "3    84.21\n",
       "4     0.88\n",
       "Name: prompt_2, dtype: float64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df.loc[df['year']==2021]['prompt_2'].value_counts(normalize = True).sort_index()*100).round(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e87f3768",
   "metadata": {},
   "source": [
    "### Method, controling for for dataset type"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "3d2d43e9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "113"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#limits to only\n",
    "len((df.loc[(df['prompt_2']==\"3\")]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "6d281bb5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2021    96\n",
       "2020    17\n",
       "Name: year, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#out of which ( , )\n",
    "df.loc[(df['prompt_2']==\"3\")]['year'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "2fc88ce6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "B    58.82\n",
       "C    35.29\n",
       "D     5.88\n",
       "Name: prompt_1, dtype: float64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df.loc[(df.year==2020) & (df['prompt_2']==\"3\")]['prompt_1'].value_counts(normalize = True).sort_index()*100).round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "630b8922",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "A    21.88\n",
       "B    39.58\n",
       "C    37.50\n",
       "D     1.04\n",
       "Name: prompt_1, dtype: float64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df.loc[(df.year==2021) & (df['prompt_2']==\"3\")]['prompt_1'].value_counts(normalize = True).sort_index()*100).round(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a42c912d",
   "metadata": {},
   "source": [
    "### Scientific, controlling for dataset type"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "f8dabd49",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NO     64.71\n",
       "YES    35.29\n",
       "Name: prompt_3, dtype: float64"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df.loc[(df.year==2020) & (df['prompt_2']==\"3\")]['prompt_3'].value_counts(normalize = True).sort_index()*100).round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "6c1aa47b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NO     94.79\n",
       "YES     5.21\n",
       "Name: prompt_3, dtype: float64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df.loc[(df.year==2021) & (df['prompt_2']==\"3\")]['prompt_3'].value_counts(normalize = True).sort_index()*100).round(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eefc460d",
   "metadata": {},
   "source": [
    "### A) Methods: version 2 (puts the overall prevalence into perspective). Robust to \"select all that aply prompting\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "2412cdee",
   "metadata": {},
   "outputs": [],
   "source": [
    "def process_list(x):\n",
    "    if ',' in x:\n",
    "        split = ((x.split(',')))\n",
    "    elif '\\n' in x:\n",
    "        split = ((x.split('\\n')))\n",
    "    elif ' ' in x:\n",
    "        split = ((x.split(' ')))\n",
    "    else:\n",
    "        split = np.nan\n",
    "    \n",
    "    if len(split)==4:\n",
    "        split = [x.strip('A) ').strip('B) ').strip('C) ').strip('D) ').strip(']').strip('[') for x in split]\n",
    "        return split\n",
    "    else:\n",
    "        return [np.nan,np.nan,np.nan,np.nan]\n",
    "\n",
    "df['has_A'] = df['prompt_4'].apply(lambda x: process_list(x)).apply(lambda x: x[0])\n",
    "df['has_B'] = df['prompt_4'].apply(lambda x: process_list(x)).apply(lambda x: x[1])\n",
    "df['has_C'] = df['prompt_4'].apply(lambda x: process_list(x)).apply(lambda x: x[2])\n",
    "df['has_D'] = df['prompt_4'].apply(lambda x: process_list(x)).apply(lambda x: x[3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "4acd79ce",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Has A\n",
      "2020: 74.77\n",
      "2021: 97.37\n",
      "Has B\n",
      "2020: 63.96\n",
      "2021: 66.67\n",
      "Has C\n",
      "2020: 72.07\n",
      "2021: 80.7\n",
      "Has D\n",
      "2020: 26.13\n",
      "2021: 19.3\n"
     ]
    }
   ],
   "source": [
    "print('Has A')\n",
    "print('2020:',(df.loc[df.year==2020]['has_A'].value_counts(normalize = True).sort_index()[1]*100).round(2))\n",
    "print('2021:',(df.loc[df.year==2021]['has_A'].value_counts(normalize = True).sort_index()[1]*100).round(2))\n",
    "\n",
    "print('Has B')\n",
    "print('2020:',(df.loc[df.year==2020]['has_B'].value_counts(normalize = True).sort_index()[1]*100).round(2))\n",
    "print('2021:',(df.loc[df.year==2021]['has_B'].value_counts(normalize = True).sort_index()[1]*100).round(2))\n",
    "\n",
    "print('Has C')\n",
    "print('2020:',(df.loc[df.year==2020]['has_C'].value_counts(normalize = True).sort_index()[1]*100).round(2))\n",
    "print('2021:',(df.loc[df.year==2021]['has_C'].value_counts(normalize = True).sort_index()[1]*100).round(2))\n",
    "\n",
    "print('Has D')\n",
    "print('2020:',(df.loc[df.year==2020]['has_D'].value_counts(normalize = True).sort_index()[1]*100).round(2))\n",
    "print('2021:',(df.loc[df.year==2021]['has_D'].value_counts(normalize = True).sort_index()[1]*100).round(2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "df1d9fed",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "24.16682376020674 8.834104964163789e-07\n",
      "0.18138232423946726 0.6701879744902434\n",
      "2.325924686840139 0.12723459245448984\n",
      "1.4957628708029054 0.22132446620094506\n"
     ]
    }
   ],
   "source": [
    "for col in ['A','B','C','D']:\n",
    "\n",
    "    a = df.loc[df.year==2020]['has_'+col].value_counts().sort_index()[1]\n",
    "    b = df.loc[df.year==2020]['has_'+col].value_counts().sort_index()[0]\n",
    "\n",
    "    c = df.loc[df.year==2021]['has_'+col].value_counts().sort_index()[1]\n",
    "    d = df.loc[df.year==2021]['has_'+col].value_counts().sort_index()[0]\n",
    "\n",
    "    test = chi2_contingency([[a,b],[c,d]],correction=False)\n",
    "    print(test[0],test[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "44158e32",
   "metadata": {},
   "source": [
    "### A) Methods: Version 3 (robust to no examples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "d1359f2a",
   "metadata": {},
   "outputs": [],
   "source": [
    "def clean_methods(x):\n",
    "    x = x.split(')')[0]\n",
    "    if x==\"A\" or x==\"B\" or x==\"C\" or x==\"D\":\n",
    "        return x\n",
    "    else:\n",
    "        return np.nan\n",
    "\n",
    "df['prompt_5'] = df['prompt_5'].apply(lambda x: clean_methods(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "506c2b29",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "A     6.96\n",
       "B    42.61\n",
       "C    42.61\n",
       "D     7.83\n",
       "Name: prompt_5, dtype: float64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df.loc[df.year==2020]['prompt_5'].value_counts(normalize = True).sort_index()*100).round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "01d4dd68",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "A    22.12\n",
       "B    46.90\n",
       "C    30.09\n",
       "D     0.88\n",
       "Name: prompt_5, dtype: float64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df.loc[df.year==2021]['prompt_5'].value_counts(normalize = True).sort_index()*100).round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "c58fdf37",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10.406468611298623 0.0012557465385720494\n",
      "0.3493529884010158 0.554479616319884\n",
      "4.049019525643402 0.04419700777793259\n",
      "6.619503707651101 0.010086793764008728\n"
     ]
    }
   ],
   "source": [
    "for col in ['A','B','C','D']:\n",
    "    a = df.loc[df.year==2020]['prompt_5'].value_counts().sort_index()[col]\n",
    "    b = len(df.loc[df.year==2020]) - a\n",
    "\n",
    "    c = df.loc[df.year==2021]['prompt_5'].value_counts().sort_index()[col]\n",
    "    d = len(df.loc[df.year==2021]) - c\n",
    "    test = chi2_contingency([[a,b],[c,d]],correction=False)\n",
    "    print(test[0],test[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3a9364cd",
   "metadata": {},
   "source": [
    "### B) Scientific impact"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "b63dc981",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NO     0.947368\n",
       "YES    0.052632\n",
       "Name: prompt_3, dtype: float64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc[df.year==2021]['prompt_3'].value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "cbf92ac6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NO     0.843478\n",
       "YES    0.156522\n",
       "Name: prompt_3, dtype: float64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc[df.year==2020]['prompt_3'].value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "edf59afa",
   "metadata": {},
   "outputs": [],
   "source": [
    "a = df.loc[df.year==2021]['prompt_3'].value_counts()[0]\n",
    "b = df.loc[df.year==2021]['prompt_3'].value_counts()[1]\n",
    "\n",
    "c = df.loc[df.year==2020]['prompt_3'].value_counts()[0]\n",
    "d = df.loc[df.year==2020]['prompt_3'].value_counts()[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "ec2f50db",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6.586002679019923 0.01027837060241125\n"
     ]
    }
   ],
   "source": [
    "test = chi2_contingency([[a,b],[c,d]],correction=False)\n",
    "print(test[0],test[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b285451e",
   "metadata": {},
   "source": [
    "### C) Open ended"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "eb57653d",
   "metadata": {},
   "outputs": [],
   "source": [
    "topic_terms = [\"predictive\",\"multidimensional\"]\n",
    "def process_adjectives(x):\n",
    "    return [i.lower().strip() for i in x.split(',')]\n",
    "\n",
    "df['prompt_6'] = df['prompt_6'].apply(lambda x: process_adjectives(x))\n",
    "\n",
    "bag_2020 = df.loc[df.year==2020]['prompt_6'].to_list()\n",
    "bag_2020 = ([x for xs in bag_2020 for x in xs])\n",
    "\n",
    "bag_2021 = df.loc[df.year==2021]['prompt_6'].to_list()\n",
    "bag_2021 = ([x for xs in bag_2021 for x in xs])\n",
    "\n",
    "index1 = pd.Series(bag_2020).value_counts(normalize = True).index\n",
    "index2 = pd.Series(bag_2021).value_counts(normalize = True).index\n",
    "overlapping_terms = set(index1).intersection(set(index2)) \n",
    "\n",
    "list_words = []\n",
    "\n",
    "for term in overlapping_terms:\n",
    "    entry = {}\n",
    "    if term in topic_terms:\n",
    "        continue\n",
    "    entry['term'] = term\n",
    "    \n",
    "    a = sum(pd.Series(bag_2020)==term)\n",
    "    b = sum(pd.Series(bag_2020)!=term)\n",
    "    \n",
    "    c = sum(pd.Series(bag_2021)==term)\n",
    "    d = sum(pd.Series(bag_2021)!=term)\n",
    "    \n",
    "    test = chi2_contingency([[a,b],[c,d]],\n",
    "                        correction=False)\n",
    "    entry['t'] = test[0]\n",
    "    entry['frac_2020'] = a/(a+b)\n",
    "    entry['frac_2021'] = c/(c+d)\n",
    "    entry['p_val'] = test[1]\n",
    "    list_words.append(entry)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "ed5f0c98",
   "metadata": {},
   "outputs": [],
   "source": [
    "words = pd.DataFrame(list_words)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "aea4da31",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>term</th>\n",
       "      <th>t</th>\n",
       "      <th>frac_2020</th>\n",
       "      <th>frac_2021</th>\n",
       "      <th>p_val</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>practical</td>\n",
       "      <td>6.386217</td>\n",
       "      <td>0.015652</td>\n",
       "      <td>0.001754</td>\n",
       "      <td>0.011501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>insightful</td>\n",
       "      <td>15.131853</td>\n",
       "      <td>0.060870</td>\n",
       "      <td>0.128070</td>\n",
       "      <td>0.000100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>methodical</td>\n",
       "      <td>5.067976</td>\n",
       "      <td>0.088696</td>\n",
       "      <td>0.054386</td>\n",
       "      <td>0.024372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>comprehensive</td>\n",
       "      <td>4.729667</td>\n",
       "      <td>0.133913</td>\n",
       "      <td>0.180702</td>\n",
       "      <td>0.029647</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             term          t  frac_2020  frac_2021     p_val\n",
       "26      practical   6.386217   0.015652   0.001754  0.011501\n",
       "8      insightful  15.131853   0.060870   0.128070  0.000100\n",
       "11     methodical   5.067976   0.088696   0.054386  0.024372\n",
       "6   comprehensive   4.729667   0.133913   0.180702  0.029647"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "words.loc[words.p_val< 0.05].sort_values(by = 'frac_2020')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "ef540791",
   "metadata": {},
   "outputs": [],
   "source": [
    "words = pd.DataFrame(list_words)\n",
    "words['frac_2020'] = (words['frac_2020']*100).round(2).astype(str).apply(lambda x: x+\"%\") \n",
    "words['frac_2021'] = (words['frac_2021']*100).round(2).astype(str).apply(lambda x: x+\"%\") \n",
    "words['t'] = (words['t']*100).round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "86088b25",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.sample(20, random_state = 0)[['text','prompt_1','prompt_3']].to_csv('annotation_sample.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "0768ded7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lllrr}\n",
      "\\toprule\n",
      "             term & frac\\_2020 & frac\\_2021 &       t &    p\\_val \\\\\n",
      "\\midrule\n",
      "       insightful &     6.09\\% &    12.81\\% & 1513.19 & 0.000100 \\\\\n",
      "        practical &     1.57\\% &     0.18\\% &  638.62 & 0.011501 \\\\\n",
      "       methodical &     8.87\\% &     5.44\\% &  506.80 & 0.024372 \\\\\n",
      "    comprehensive &    13.39\\% &    18.07\\% &  472.97 & 0.029647 \\\\\n",
      "         detailed &     2.43\\% &     3.86\\% &  190.84 & 0.167144 \\\\\n",
      "      informative &     0.87\\% &     1.58\\% &  119.27 & 0.274793 \\\\\n",
      "      inquisitive &     0.17\\% &     0.53\\% &  102.11 & 0.312251 \\\\\n",
      "           robust &     0.52\\% &     0.18\\% &   98.61 & 0.320704 \\\\\n",
      "   methodological &     0.52\\% &     0.18\\% &   98.61 & 0.320704 \\\\\n",
      "     quantitative &     0.52\\% &     0.18\\% &   98.61 & 0.320704 \\\\\n",
      "    collaborative &     3.13\\% &     2.28\\% &   78.46 & 0.375728 \\\\\n",
      "        impactful &     9.91\\% &     11.4\\% &   66.80 & 0.413761 \\\\\n",
      "       innovative &    14.78\\% &    13.16\\% &   62.86 & 0.427885 \\\\\n",
      "         relevant &     1.57\\% &     2.11\\% &   46.37 & 0.495906 \\\\\n",
      "        strategic &     0.17\\% &     0.35\\% &   34.30 & 0.558086 \\\\\n",
      "       systematic &     0.17\\% &     0.35\\% &   34.30 & 0.558086 \\\\\n",
      "        inclusive &     0.35\\% &     0.18\\% &   32.55 & 0.568311 \\\\\n",
      "multidisciplinary &     0.35\\% &     0.18\\% &   32.55 & 0.568311 \\\\\n",
      "      resourceful &     0.35\\% &     0.18\\% &   32.55 & 0.568311 \\\\\n",
      "         in-depth &     0.52\\% &     0.35\\% &   19.22 & 0.661088 \\\\\n",
      "         rigorous &     1.04\\% &     0.88\\% &    8.32 & 0.773025 \\\\\n",
      "         thorough &     1.91\\% &     2.11\\% &    5.37 & 0.816694 \\\\\n",
      "       analytical &    14.09\\% &    14.39\\% &    2.10 & 0.884885 \\\\\n",
      "      data-driven &     1.57\\% &     1.58\\% &    0.03 & 0.985101 \\\\\n",
      "       meaningful &     0.17\\% &     0.18\\% &    0.00 & 0.995068 \\\\\n",
      "  detail-oriented &     0.17\\% &     0.18\\% &    0.00 & 0.995068 \\\\\n",
      " forward-thinking &     0.17\\% &     0.18\\% &    0.00 & 0.995068 \\\\\n",
      "           timely &     0.17\\% &     0.18\\% &    0.00 & 0.995068 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/m5/8d669b8n0vl1_87nc6mbcytm0000gn/T/ipykernel_50911/1375450471.py:1: FutureWarning: In future versions `DataFrame.to_latex` is expected to utilise the base implementation of `Styler.to_latex` for formatting and rendering. The arguments signature may therefore change. It is recommended instead to use `DataFrame.style.to_latex` which also contains additional functionality.\n",
      "  print(words[['term', 'frac_2020', 'frac_2021','t', 'p_val']].\\\n"
     ]
    }
   ],
   "source": [
    "print(words[['term', 'frac_2020', 'frac_2021','t', 'p_val']].\\\n",
    "sort_values(by = 't', ascending = False).to_latex(index=False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e3fe0f13",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}
